Processing WES VCF for case control GWAS analysis
Entering edit mode
4 months ago

Dear community

I'm currently trying to try a GWAS for case control study using PLINK using whole exome VCF files. I understand that I need to convert these VCF files into PLINK's binary format (bed/bim/fam) before I can proceed with the GWAS.

From my literature I have to manually create phenotype and covariate data for the individuals in my study. Like this


 Sample1 Sample1 1
Sample2 Sample2 2
Sample3 Sample3 1


FID     IID     SEX
Sample1 Sample1 1
Sample2 Sample2 2
Sample3 Sample3 1


#CHROM  POS ID  REF ALT QUAL    FILTER  INFO    FORMAT  Sample1 Sample2 Sample3
chr1    14653   1/1 C   T   411/127 MG_SNP_Filter   AC=2;AF=01/15;AN=4;BaseQRankSum=11/111;ClippingRankSum=0;DP=22;FS=0;MLEAC=1;MLEAF=01/15;MQ=251/178;MQRankSum=01/137;QD=21/19;ReadPosRankSum=-11/1664;SF=0f,2f;SOR=01/1941;set=FilteredInAll GT  0/1 1/1 0/1
chr1    14677   1/1 G   A   501/177 MG_SNP_Filter   AC=1;AF=01/15;AN=2;BaseQRankSum=-21/1054;ClippingRankSum=-11/1203;DP=9;FS=131/1222;MLEAC=1;MLEAF=01/15;MQ=301/151;MQRankSum=11/12;QD=51/164;ReadPosRankSum=01/1248;SF=2f;SOR=21/178;set=FilteredInAll   GT  1/1 1/1 0/1

Then I run ./plink --vcf ~/plink/merge/analysis/head.vcf --pheno ~/plink/merge/analysis/pheno.txt --covar ~/plink/merge/analysis/covar.txt --make-bed --out ~/plink/merge/analysis/head

they generate .fam .bim .bed .cov correctly (and nosex file for some reason?)

after that I tried to run association with ./plink --bfile ~/plink/merge/analysis/head --pheno ~/plink/merge/analysis/pheno.txt --assoc --out ~/plink/merge/analysis/assoc_results

but the log file show that Note: No phenotypes present. Warning: Skipping --assoc/--model since less than two phenotypes are present.

So I have to add --allow-no-sex to done the job.

My question is why the sex is not load? I have check covariate file that wrote by plink still show the sex number.

Thank you in advance for your help!

GWAS PLINK • 415 views
Entering edit mode

I think that --covar is not meant to represent the sex of your individuals in plink. It is supposed to be a covariate baut it could be anything. To run, plink requires by default to have the sex of each individual in the .fam file. Since sex is not contained in VCF, this information is set to missing during conversion. You can either ignore it as you did (and your sex will be used as a covariate of your model) or use --update-sex ~/plink/merge/analysis/covar.txt to set your individual's sex.


Login before adding your answer.

Traffic: 1251 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6